Organisms constantly update their behavior in response to sensory stimuli and experiences. The process by which an organism alters its behavior in response to experiences is known as learning. While the parameters and rules for acquisition and storage of learned information have been well established and are quite similar across species ranging from C. elegans to mammals, the molecular mechanisms that drive learning, in comparison, are not well understood. In animals, learning is not directly measurable but can be measured by performance. One basic form of learning that is readily quantifiable is non-associate learning, which includes sensitization and habituation. Habituation is defined as a decrease in attention directed towards a stimulus. For example, the startle reflex is stereotyped but highly modifiable, such that repeated presentation of startling stimuli suppresses the startle reflex. In humans, habituation deficits are a central feature of cognitive disorders, including schizophrenia. Zebrafish larvae display a remarkable repertoire of behaviors and we find that larvae exhibit non-associative learning (habituation) with landmark behavioral and pharmacological characteristics. Moreover, we have devised an experimental setup that allows us perform genetic as well as small molecule screens for genes and pathways specifically underlying non-associative learning (habituation). The objective of this proposal is to generate a large mutant collection, thereby forming a platform for integrative studies aimed to understand the neural, molecular and clinical aspects of non-associative learning. The experiments in this proposal will: (1) screen an equivalent of 1,000 chemically mutagenized genomes, and 500 strains carrying gene-breaking insertional mutations for defects specifically in short-term habituation (through automated high-speed video analysis); (2) screen libraries of small molecules with known targets to identify factors that increase or decrease habituation to define entry points into pathways underlying non-associate learning; and (3) map and clone selected mutants (through a combined candidate/positional cloning approach). These studies are relevant to the study of human diseases, as deficits in startle modulation have been described in several mental disorders, including schizophrenia, attention deficit disorder, and obsessive-compulsive disorder, and are also are common in individuals with a history drug abuse.